RuntimeError: CUDA out of memory. Tried to allocate 250.00 MiB (GPU 0; 2.00 GiB total capacity; 540.10 MiB already allocated; 133.94 MiB free; 552.00 MiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation.  See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF

batch_size = 1 了，还是崩溃，我没办法了。谁能支援我一个高配显卡的电脑啊，我可以帮你干活。
--
FROM 120.242.251.*

就算能跑，是不是也要自己搭环境？比如安装cuda什么的？
【 在 zszqzzzf 的大作中提到: 】
: 能跑。
--
FROM 120.242.251.*